Successful Aging

The projected increase in the population of older adults (ages 65 and older) brings with it concerns about unavoidable loss through age-related chronic illnesses like dementia, cancer, or arthritis (Jaul & Barron, 2017). Successful aging is a multidimensional concept that acknowledges potential morbidity but offers a compelling alternative to unavoidable loss by incorporating dimensions of aging other than physical health (Rowe & Kahn, 1998). As successful aging is an inherently multidimensional concept it should not be limited to being explored in a unidimensional manner. However, there have been few attempts to study it comprehensively. Instead, many studies tend to examine unidimensional aspects of successful aging, particularly focusing on physical health (Cosco et al., 2014). Latent variable analysis offers a methodological approach that allows for a flexible understanding of successful aging. However, few studies have utilized such an approach (Mana & Bezdicek, 2020), leaving it ripe for further development. There is potential in utilizing the marital relationship to promote successful aging through improving relationship processes. The marital relationship, specifically, is a critical focus as it has the possibility to shape health outcomes through the distinct processes of support and strain (House et al., 1988). However, outside of a few exceptions (Ko et al., 2007; Tracy & Utz, 2020), there has been little exploration of the marital relationship processes that may affect a person's overall health and well-being and subsequent ability to age well. Based on the assumption that successful aging is a multidimensional concept, these two studies work together to better understand the multiple factors contributing to successful aging, particularly the marital processes of support and strain. As such, both papers are interrelated. Study 1 was designed to apply the LPA statistical method as a strategy to produce a comprehensive outcome for successful aging. The samples included N=2,257 participants from the MIDUS 2 and N=3,351 participants from the MIDUS Refresher. Latent profile analysis (LPA), a statistical method that allows for the grouping of individuals without using preconceived determinants, (Gibson, 1963), was applied to various indicators of the dimensions of physical, mental, and cognitive function and active engagement with life. To evaluate the criterion validity of the identified latent profiles in the MIDUS 2, the same measures were analyzed in a demographically matched cohort of adults from the MIDUS Refresher. To evaluate the content validity of the profiles, demographics including race, age, sex, ethnicity, and educational attainment were explored with the profile assignment for both MIDUS 2 and MIDUS Refresher, respectively. Results from Study 1 indicated four distinct profiles two demographically matched data sets: the MIDUS 2 and the MIDUS Refresher. These profiles were reproduced in a demographically similar cohort of adults. Finally, the demographic correlates of the profiles of successful aging do follow patterns like those observed in single dimensional assessments of health, and identified latent profiles generally comport with descriptions of health in the general population. The results from Study 1 provided an expanded understanding of successful aging by verifying the validity of the latent profiles. Study 2 used the profiles created in Study 1 and multinomial logistic regression to explore the contributing factors for spousal support and strain at time one (MIDUS 1), on the successful aging profile at time two (MIDUS 2). The goal of this study was to explore the impact of the relationship processes of marital support and strain on successful aging. The final sample included n=868 participants from the MIDUS 1 and MIDUS 2 who were married at time one and had been married only once by time two. Marital dissolution (i.e., separation, divorce, widowhood) by time two was also explored as a covariate. Potential gender differences in the association of spousal support and spousal strain with successful aging were also explored as an interaction model. Results from the multinomial logistic regression model for Study 2 indicated that spousal support at time one was not associated with less detrimental profiles of successful aging at time two, and spousal strain at time one was associated with more detrimental profiles of successful aging at time two. There was no significant relationship between spousal support, spousal strain, gender, and successful aging profiles in the interaction model. This study exemplified the potential to apply family science measures to a multidimensional successful aging outcome.